\begin{table} \small \centering \protect \captionsetup{justification=centering} \caption{\label{tab:table5} The Impact of Increased Representation on Mining Conflict Onset (Robustness) }
\noindent \scalebox{1}{ \begin{threeparttable}
\begin{tabular}{lccc}
\toprule
                    &        30km&        40km&        50km\\
                    &\multicolumn{1}{c}{(1)}&\multicolumn{1}{c}{(2)}&\multicolumn{1}{c}{(3)}\\
\hline
PESA X Scheduled    &      0.0060&      0.0101&      0.0111\\
                    &    (0.0037)&    (0.0051)&    (0.0063)\\

\midrule Mean Pre-Y (Non-Sch)&      0.0133&      0.0233&      0.0361\\
Mean Pre-Y (Sched)  &      0.0344&      0.0593&      0.0854\\
Dataset             &         VCF&         VCF&         VCF\\
Timespan            &   1995-2017&   1995-2017&   1995-2017\\
\midrule \emph{Fixed-effects}&            &            &            \\
Pixel               &$\checkmark$&$\checkmark$&$\checkmark$\\
State $\times$ Year &$\checkmark$&$\checkmark$&$\checkmark$\\
\midrule \emph{Time Trends}&            &            &            \\
t (Pixel)           &$\checkmark$&$\checkmark$&$\checkmark$\\
\midrule \emph{Fit statistics}&            &            &            \\
\# Pixel            &       30843&       30843&       30843\\
\# State $\times$ Year&         198&         198&         198\\
\# Observations     &      678546&      678546&      678546\\
Standard-Errors     &       Block&       Block&       Block\\
R^2                 &       0.812&       0.816&       0.821\\
\hline
\hline
\end{tabular}
\begin{tablenotes}
\begin{footnotesize}
\emph{Notes:} Standard errors are clustered at the block level and reported in parentheses. Conflict data is from \texttt{https://landconflictwatch.org/}. Conflict onset is assigned to all pixels within the indicated radius in each panel.
\end{footnotesize}
\end{tablenotes} \end{threeparttable} } \end{table}
